​Error Proofing and Public Policy Diffusion

There are doubtless many lessons in many different areas of public policy that will be drawn from the experience around the world of dealing with the COVID 19 crisis. This blog looks at lessons around the diffusion of public policy.

Background
 
We live in a world where change that occurs in one corner of the globe can very rapidly spread to another and, in some circumstances, become a worldwide phenomenon. In the private sector the incentive for businesses is to catch onto new developments as quickly as possible in order to gain ‘first mover’ advantages in market positioning. For the public sector, incentive signals are more mixed. A major concern is to avoid error. There are gains to be made in latching on to new policy approaches, but mistakes could be costly and, in democratic countries, will be punished at the ballot box. Calculations of the upside and downside risks of public policy innovation are different.
 
In responding to the COVID pandemic, governments faced an urgent common problem for public policy and had a strong incentive to see what other governments were doing to combat the pandemic and to copy whatever appeared to be successful. Nevertheless, what stands out is the diversity of policy responses, with wide variations in infection and mortality rates as a result.
 
In the case of COVID the diversity of responses appears to have constituted a public ‘bad’ – reflecting a costly delay in learning, or an unwillingness to learn from others, about effective responses. This blog looks more generally at the dynamics of diffusion in public policy making. Contrary to appearances, a diversity of responses is likely to be a ‘good’ because a simple ‘copy-out’ of policies elsewhere is rarely likely to be applicable. The blog looks first at the drivers of diffusion and secondly, at the protections against errors and misapplications in policy diffusion. 
 
Drivers
 
The drivers of policy diffusion run along a spectrum. They can broadly be divided into those that are ‘spontaneous’ and those that are more considered.
 
Technical Innovation 
 
One of the first studies of diffusion looked at the case of the adoption of hybrid corn in the United States in the 1930s. In this case the advantages in terms of higher yields and higher income were so persuasive that a switch to hybrid corn caught on ‘spontaneously’ through seed companies. The widespread adoption of high yielding rice provides a more recent example of science-based diffusion. The introduction of vaccines against COVID 19 fits within this category. In these cases of science-based diffusion the implication for public policy is about support – farm extension services in the case of hybrid corn, water management policies in the case of rice, and vaccine approval and distribution in the case of COVID.
 
Policy Emulation 
 
A second driver of diffusion is a simple desire to emulate a policy that seems to be successful elsewhere. In this case the source of innovation is not science based as in the case of hybrid corn, or COVID vaccines, but policy based. 
 
An example of widespread policy emulation is provided by the diffusion of what came to be called the ‘New public management’ in the 1980/1990s. Starting with the example of public sector reforms in New Zealand under Roger Douglas (Finance Minister, 1984-88) the mantra that governments should ‘steer but not row’, became a guiding light for public sector reforms elsewhere. Outsourcing and ‘agencification’ of functions previously centralised within public administrations became an almost world-wide phenomenon. 
 
Benchmarking
 
Policy emulation is sometimes triggered by the increasing use of international benchmarks to measure public policy outcomes. One benchmark with widespread influence has been the World Bank’s ‘Ease of Doing Business’ indicators which place Singapore and New Zealand at one end and countries such as Chad and the Central African Republic at the other. In this case, governments feel that they will be more successful in attracting foreign investment if they rank high on the table. Another influential benchmark is OECD’s international comparison for high schools (the Program for International Student Assessments, PISA). In the case of education, governments pay attention to outcomes elsewhere because spending on public education accounts for around 3.5 % of GDP across the OECD area while the results often seem disappointing. Benchmarking has become an independent driver of policy diffusion.
 
What stands in the way of ‘spontaneous’ diffusion, driven by scientific or policy innovation, or benchmarking, and diffusion motivated by a simple desire to emulate others, is the importance of context. For example, the adoption of hybrid rice has involved related issues of property rights and water management. The prescriptions of the new public management had to be fitted into different public sector, civil service and administrative traditions. Even carefully constructed benchmarks require translation into actual policy. Thus, there is a need for more considered approaches to policy diffusion that pick up and select the correct lessons. 
 
Considered diffusion
 
Peer review
 
One of the main techniques for a considered diffusion of public policies and for lesson learning is peer review. Peer review allows for more in-depth comparisons that take into account country specifics. OECD is one of the traditional leaders in this field. For example, it monitors the implementation of its Anti-Bribery Convention through compulsory peer reviews of the policies and practices of the 40 + signatories to the Convention.
 
Best practice 
 
Lesson learning is sometimes encapsulated in principles of ‘Best practice’. An example is provided by OECD’s ‘Best Practice Principles for Regulatory Policy and Governance’ that in turn serve as a basis for peer review. 
 
Coercion
 
The diffusion of even well considered public polices is not always an entirely voluntary process. When borrowers accept policy conditions from the IMF, in return for financial assistance and for help in negotiating debt relief with other lenders, the policies may follow fairly well-trodden pathways of what has worked elsewhere. Nevertheless, adoption reflects a setting where the borrowing country has little choice. More broadly, international agreements to adopt common approaches may conceal what are called ‘disaffected minorities’ who accept common policies unwillingly because of pressures on them to contribute to a consensus. 
 
Error proofing
 
There are two main sources of error in the transfer of public policy. The first concerns again the importance of context. Even where policy diffusion is a considered process, what works in one context may not work when transferred to the particularities of another. Secondly, is the importance of public opinion in democracies. Policy needs to be ‘evidence based’, but it also has to take into account value judgements as reflected in political preferences. These may well vary between jurisdictions.
 
There are three main classes of technique to try to avoid errors in diffusion. The first centres on processes. The second is about ensuring ‘voice’. The third is about parliamentary scrutiny.
 
Processes
 
The main defence against making mistakes about the transferability of models is to express policy transfers in terms of principles rather than precise rules. The distinction places the emphasis on the equivalent effect of policies rather than on equivalent means and instruments. In the EU the distinction is recognised in the difference between an EU directive that embodies principles and a directly applicable EU regulation that conveys rules. 
 
Voices
A second defence concerns the need to include different voices in policy making. The key class of actors in policy transfers are the professional practitioners in the field in question – the financial regulators, the health, education and other professionals. Procedures typically allow for other interests that are directly affected by a policy transfer to have the opportunity for comment. Private interests may also be brought into the assessment in other ways, such as through advisory committees. 
 
The difficult area is in relation to public opinion more generally. Private sector NGOs and government advisory committees are not necessarily representative of public opinion as a whole but may be advocates of a particular ethical position or interest not shared in the same way by the general public. So-called ‘civil society’ organisations may be selected to boost the views of the administration.
 
Scrutiny
 
Public opinion can be consulted directly on policy transfers through such techniques as ‘Town Hall’ or ‘Citizen Assembly’ gatherings. However, it is elected representatives who generally have to stand in for the public in making judgments about what is acceptable. When the policy transfer requires enacting domestic legislation, then normal scrutiny procedures by elected representatives are likely to apply. What may slip through the net are those transfers that require only administrative or agency action.
 
Conclusions
 
Governments around the world face many common problems - from underperforming systems of public education, to difficulties in dealing with aging populations. There are gains to be reaped from looking at the successes and failures of other governments. However, policy diffusion is error prone and error proofing is required. Error proofing slows diffusion. Thus, delays in emulating others may be for good reason as well as for bad. However, the mixed record in COVID suggests that procedures for international lesson learning remain far from optimal in helping different jurisdictions to strike the right balance between the benefits of speed and the costs of mistakes. The reputation of WHO has been badly damaged and at some point its operations will need to be reviewed.